We propose to continue in vitro studies of the kinetics of formation of native three-dimensional structure in proteins, to compare various conditions for rapid structure formation, and to obtain information on the mechanism(s) of the self-assembly process with emphasis on physiological relevance. The identity and distribution of disulfide cross-linked intermediates will be determined at various stages of the regeneration of reduced, disorganized proteins. Limited proteolysis followed by column chromatographic separations make it possible to separate disulfide-containing peptides, which will be identified by amino acid composition. The time-pattern of formation of disulfides outlines a folding pathway or a manifold of such pathways. In some cases additional information on the folding pathway may be obtained from parallel kinetic study of the circular dichroic spectra and the fluorescence emission spectra. We will test the ability of fragments of proteins to form native-like structures from unordered structures. Native-like structures will be evaluated by disulfide-bonding pattern, catalytic and/or binding functions, and immunochemical cross-reactivity. In cases where fragments do fold to native-like structures, kinetic studies of the folding will be carried out. We will develop theoretical approaches to the analysis of globular protein structure, using geometric constraints for an early reduction of computational complexity. We will work in the framework of viewing protein structure as a hierarchial assembly of structural modules. This can lead to modeling three-dimensional structure by construction.